Usefulness of an S-1 dosage formula: an exploratory analysis of randomized clinical trial (JCOG1001)

Gastric cancer is the fourth leading cause of cancer-related deaths worldwide [1] and is the third leading cause of cancer-related deaths in Japan [2]. For locally advanced resectable gastric cancer, perioperative chemotherapy and curative surgery are recognized as standard treatments because of the high recurrence rate in surgery alone. In Asia, postoperative adjuvant chemotherapy after a curative resection is the standard of care. The ACTS–GC trial showed a survival benefit of adjuvant S-1 monotherapy for 1 year after the curative resection of pathological stage II/III gastric cancer [3]. The JACCRO-GC-07 trial showed a significant prolongation of overall survival (OS) using adjuvant chemotherapy with S-1 plus docetaxel compared with S-1 monotherapy for patients with pathological stage III gastric cancer [4]. Several studies demonstrated that the completion of a planned adjuvant S-1 therapy is associated with favorable survival outcomes of patients with gastric cancer after a curative resection [5‐7]. However, when severe adverse events occur, the treatment is interrupted and terminated. To achieve completion, an appropriate dose reduction and/or treatment schedule alteration are required in some patients.

S-1 is a pro-drug of 5-fluorouracil (5-FU), consisting of tegafur, gimeracil (CDHP), which is an inhibitor of dihydropyrimidine dehydrogenase catabolizing 5-FU, and oteracil. Since more than 50% of CDHP is excreted by the kidneys, its excretion is reduced in patients with low renal function, resulting in persistently high blood 5-FU concentration [8‐10] and high frequency of adverse events. Therefore, it is recommended to reduce its dose in cases with low renal function [11]. A low renal function was considered as one of the risk factors for poor compliance of adjuvant chemotherapy with S-1 [12]. In contrast, for cases with good renal function and large body surface area (BSA), the highest approved S-1 dose (120 mg/day) may not be sufficient to achieve the target AUC, and its efficacy may be reduced. It is also important to determine the appropriate S-1 dose for such patients. However, there is no established method, such as the Calvert formula for carboplatin, for patients’ renal function. Therefore, to obtain the target pharmacokinetics of S-1, the S-1 dose is determined according to BSA and is reduced based on the clinical experiences of patients in the clinical practice.

Booka et al. developed a formula for determining the S-1 dosage based on renal function [13]. The validation of this formula led to the development of a revised formula for the determination of the S-1 dosage, which considered the gender of the patient. The revised formula is believed to be useful in metastatic settings [14]; however, it has not been validated in adjuvant settings.

The Japan Clinical Oncology Group (JCOG) study JCOG1001 is a randomized phase 3 trial that investigated the role of bursectomy for cT3/T4 resectable gastric cancer [15], which did not show survival benefits. The enrolled patients received adjuvant S-1 therapy for 1 year as a protocol treatment. The data regarding the clinical courses during adjuvant chemotherapy was collected. This study aimed to explore the clinical utility of the S-1 dosage formula in adjuvant chemotherapy with S-1 for gastric cancer using the data of JCOG1001.

In summary, this study demonstrates that the OD group had the highest frequency of grade ≥ 3 adverse events and the lowest proportion of completion of adjuvant chemotherapy with S-1 for 1 year, but that there was no difference in RFS and OS among the three groups.

Although it was expected that the proportion of completion of adjuvant S-1 would be the highest in the UD group and the lowest in the OD group, the proportion of completion in the UD group (74.1%) and ED group (73.7%) was similar, and it was the lowest in the OD group (68.5%). Dose reduction and/or treatment schedule alteration are usually performed after the occurrence of grade ≥ 3 adverse events. Because appropriate management also had a positive impact on completion, the impact of dose optimization by the S-1 formula dosage might be reduced in this study. However, considering that severe adverse events were most frequently observed in the OD group at the early phase of adjuvant therapy, which required dose reduction or treatment schedule alteration, and that there was no significant difference in the OS and RFS of the three groups, dose optimization using this formula may be useful to reduce the toxicities in patients classified as the OD group without interfering with its efficacy in clinical practice.

In the palliative setting of S-1 monotherapy (SPIRITS trial [16]), Booka et al. compared the survival outcomes of patients with stage IV gastric cancer classified into the three dose groups by the S-1 dosage formula. It was reported that the OS in the ED group was longer than that in the OD and UD groups [17]. In this study, reproducibility was expected in adjuvant settings. However, there were no significant differences among the three groups, and the survival of the UD group was longer than that of the other groups. There are some reasons to be considered for this discrepancy. First, it is probably due to patient background differences. More patients, who were younger, male, and had higher BMI, were considered to have a better long-term prognosis independently from the efficacy of adjuvant chemotherapy [3, 18]; they were classified into the UD group due to the nature of the formula. Second, patients often experience body weight loss or lowering renal function during adjuvant treatment. As a result, patients who were classified into the UD group might be switched to the ED group because of the body weight loss during treatment. Therefore, it is considered that dose optimization by this formula may be necessary to be repeated according to the patient’s condition. In contrast, while grade ≥ 3 adverse events were observed most frequently in the OD group, the dose of S-1 in the OD group was reduced once adverse events occurred. It was speculated that dose reduction could adjust the S-1 dose in the early period after initiating the adjuvant chemotherapy in the OD group, resulting in treatment completion. Moreover, schedule alteration could reduce the toxicities and maintain the dose intensity regardless of the S-1 dose. Schedule alteration might also reduce the impact of S-1 dose optimization using the S-1 formula on the survival outcomes. The S-1 dosage formula was originally developed for patients with stage IV gastric cancer, and it is unclear whether the formula can be sufficiently extrapolated to adjuvant settings. In this study, even patients in the ED group were required to reduce dosage or skip treatment more frequently compared with patients with advanced gastric cancer who were treated with S-1 monotherapy in the ED group of the SPIRITS trial. The pharmacokinetics of S-1 showed no clinically significant difference immediately before and after gastrectomy [18]. Moreover, during the development of the S-1 formula, gastrectomy showed no relevant impact on the pharmacokinetics of S-1 in patients with unresectable and recurrent gastric cancer [13]. Therefore, it is possible to pharmacokinetically extrapolate this formula to the gastrectomized patients. However, as opposed to the results of the SPIRITS trial, the patients in this study often experienced body weight loss after gastrectomy as described above. Hence, the maintenance of body weight through nutrition support and application of this formula, which is based on the on-time body weight following gastrectomy, might be helpful in reducing toxicities caused by S-1 in adjuvant settings.

At present, the treatment guidelines for gastric cancer recommend S-1 plus docetaxel or oxaliplatin combination therapy, with S-1 or capecitabine as postoperative adjuvant therapy for patients with stage III gastric cancer. The frequency of adverse events tends to be higher in doublet therapy containing S-1 than in S-1 monotherapy. In the aforementioned study on the usefulness of the S-1 formula in patients with stage IV gastric cancer in the G-SOX trial [17], the frequency of adverse events tended to be higher in the OD group than in the ED group. In particular, neutropenia, anemia, and creatinine levels were significantly higher in the OD group than in the ED group. As the dose optimization of S-1 using the S-1 formula may potentially reduce the adverse events of doublet chemotherapy in patients with stage IV gastric cancer, it is expected that dose optimization using this formula is useful for reducing the toxicities of doublet chemotherapy in adjuvant settings.

This study has several limitations. First, body weight often changes during adjuvant treatment; therefore, the formula might not work accurately. Second, the patients’ backgrounds were not balanced among the three groups. As described above, patients who were younger, males, and had higher BMIs tended to be classified in the UD group, whereas those who were older, females, and had lower BMIs tended to be classified in the OD and UD groups due to the nature of the formula. It is believed that these differences in the patients’ background characteristics have some impact on the clinical outcomes. Although it is difficult to adequately avoid this bias, the hazard ratios adjusted for patients’ background characteristics were also calculated in the multivariate analysis for survival. Considering the similar survival among the three groups in the multivariate analysis, it is suggested that the dose optimization by the S-1 formula, especially for the OD group, might reduce the toxicities without interfering with OS and RFS. Finally, because this formula was only developed for Japanese patients, the results of this study cannot be extrapolated to patients in Europe and the United States who show different pharmacodynamics of S-1 from Asian patients.

In conclusion, dose optimization using the S-1 dosage formula may potentially reduce grade ≥ 3 adverse events for the elderly, lower BMI, or female patients who may be overdosed. For adjuvant settings, it may be recommended to use this S-1 dosage formula, considering the chronological changes in body weight and renal function.